Medical Error Is Not the Third Leading Cause of Death

by Brian Dunning
August 6, 2024

It's one of the Internet's favorite perennial false claims, and it's so persistent that it remains today a current political talking point. It concerns medical errors, which are things like patients being given drugs they're allergic to; bedsores resulting in infection; healthcare-acquired infections; failure to spot things like sepsis; and misdiagnoses or wrong treatments or failure to follow treatment plans. The claim that medical errors like these are the third leading cause of death in the United States is one that you can find plastered just about anywhere you look for it, even though it was never true, and not even close to true. Volumes of papers have been published showing it to be false — and not just false, but false by more than an order of magnitude. Today we're going to find out where this urban legend came from, reveal how we know it's not true, and see why the heck so many people continue to repeat it.

The "why" is easy to answer. The idea that medical errors in hospitals kill an enormous number of people is one that fits nicely into the agendas of many groups. For example, the National Rifle Association uses this talking point to claim that "medical misadventures," as they term it, kills more people than guns. Far right pundit Ben Shapiro has repeated this falsehood to bolster another of his favorite falsehoods that "diversity hires" at hospitals have made medical care unsafe because doctors have been hired for their ethnicity rather than for their skills. Shapiro has made this claim both on Twitter/X and in his video "DEI Could Get You Killed in the Operating Room." Medical malpractice law firms and conservative lawmakers urging tort reform all commonly repeat the trope; one such law firm's website stating:

According to a recent study from John[sic] Hopkins, at least 250,000 Americans die as a result of medical error every year. Some estimates indicate the number of Americans who die due to medical mistakes may be close to twice that number.

Even if we rely on conservative estimates, these findings indicate that medical error is the third-leading cause of death in the US. Only cancer and heart attacks claim more lives every year.

So if this is untrue — and as we're going to find out, it's extremely untrue — how did this misinformation get started and spread so widely?

In 1999, the Institute of Medicine (today the National Academy of Medicine), part of the US National Academies, commissioned a report looking for ways to improve the quality of healthcare in the United States. The result was a 300-page book, To Err Is Human: Building a Safer Health System, edited by Kohn et al. As one might expect, it did not receive a tremendous amount of fanfare in the press, but it did receive a tidal wave of academic criticism. Its second chapter, "Errors in Health Care: A Leading Cause of Death and Injury" made the case that at least 44,000 Americans, and possibly as many as 98,000, die from preventable medical errors in hospitals each year. This study was the source of Health and Human Services secretary Xavier Becerra's plane crash analogy when he stated in 2022 that "We're losing pretty much an airline full of Americans every day to medical error." How did they come up with those numbers?

They came from two small local studies, which the authors then extrapolated to the population size of the US. The problems with that are several, most often cited by critics being that the patient populations studied (mostly people with less than 3-month life expectancy) were not representative of the United States as a whole; and that the connections between the errors and the deaths were not necessarily causal. Most medical errors do not cause death. So their final post-extrapolation numbers were wildly exaggerated.

But even if that higher 98,000 number was correct, that still wouldn't make medical errors the third leading cause of death in the United States; not even close. In 2022 (the last year for which final numbers are available as of this writing), the leading causes (in order) were heart disease (703,000), cancer (608,000), followed by accidents, COVID-19, stroke, lower respiratory diseases, Alzheimer's, diabetes (at 101,000), and then the 98,000 number would rank only ninth. So then where does this "third leading cause of death" claim come from?

That comes from a now-infamous 2016 article published in the BMJ by Martin Makary and Michael Daniel, and it's the Johns Hopkins paper mentioned earlier. Its title leaves little room for ambiguous interpretation: "Medical error — the third leading cause of death in the US" and, with the Johns Hopkins press office behind it, it garnered worldwide headlines. This paper is the culprit behind one of today's most abused pieces of misinformation.

Let's look at its problems. First of all, it's very important to point out that this article was only commentary, there was no scientific research behind it, and (crucially) it was not peer reviewed. What Makary and Daniel did was to look at only four papers that studied a total of only 35 deaths due to preventable adverse events. They simply used the number of total hospital admissions in 2013 to multiply that 35 into 251,454 annual deaths by medical error, which would indeed put medical errors third on the list behind heart disease and cancer. Obviously, the extrapolation of such a huge number from such a tiny sample wouldn't satisfy any statistical scrutiny.

What the Johns Hopkins press office did not advertise, and which apparently no reporters on the planet read, were the response articles published directly alongside this paper. Here are some of the problems noted:

1. Approximately 700,000 deaths happen in hospitals each year. For 251,000 of them (over a third) to have been preventable and caused by medical error is wildly implausible.
   
   
2. The authors neither proposed nor used any formal methodology, especially problematic since the four studies they looked at used different methodologies and different definitions for what they considered medical errors.
   
   
3. The authors called for certain robust studies to be performed. A number already had been, all of which found far lower numbers.
   
   
4. Many of the responses were critical of the article's attention-grabbing and sensationalist headline, calling the authors' motivations into question.

Even the BMJ's own editors of BMJ Quality and Safety laid harsh criticism on Makary and Daniel. Their criticism was too long even to effectively summarize here, but it included criticism of the four studies upon which they based their commentary; their lack of methodology; the invalidity of the extrapolation of huge numbers from small ones; and the causal problem where dying with a condition is not the same thing as dying from that condition.

Despite all the criticism, including their own, the BMJ has stated they have no plans to retract Makary and Daniels' paper, on the principle that it was commentary only and was presented as such, and contained no factual errors.

Plenty of studies have been done to find out what the true numbers are, and the vast majority of them are in agreement within a few percentage points. Perhaps the paper most often cited, and best representative of the consensus, comes from researchers at Yale and was published in 2020 in the Journal of General Internal Medicine. The paper, "Rate of Preventable Mortality in Hospitalized Patients: a Systematic Review and Meta-analysis," was by Rodwin et al. and looked at sixteen large scale studies of preventable hospital deaths in the United States. Eight met their criteria for highest quality, and it should be noted that none of the four studies used by Makary and Daniels qualified. These studies looked at a total of 12,503 deaths, not just 35, so it was a far more useful sample. There were two very important findings.

• First of all, the total rate of preventable mortality was found to be 3.1%, not the thirty-something percent Makary and Daniels claimed. This equates to about 22,165 preventable deaths annually in the United States.
  
  
• Second, the vast majority of those deaths were among patients with a life expectancy of less than three months. That means these were very sick people who were on their way out anyhow, and whatever the medical error was only sped the process, due to their weakened state. When it comes to patients with a life expectancy over three months, these people are generally less ill and are more resilient. In their case, fewer than 1% died from medical error. More simply stated, although 3.1% of hospital deaths are caused by medical errors, it's less than 1% for people who aren't already terminally ill. The vast majority of medical errors are not lethal, especially not for people who aren't dying anyway.

So the popular number on the Internet, given by Ben Shapiro and others trying to frighten you away from hospitals, is wrong by more than a factor of 11. That's more than a full order of magnitude. Medical errors are not even on the list of leading causes of death, much less third on it.

One might ask what's the harm in spreading this greatly inflated number. Isn't it better to draw more attention to the cause of patient safety? I will close with the paragraph the editors of BMJ Quality and Safety used to close their own paper:

We are deeply committed to improve patient safety and quality of care. Avoidable deaths and suffering can best be reduced by improving the evidence base and that must start with sound epidemiology. Without this, over time implausible estimates of deaths due to medical error will do more to erode the cause of patient safety than headline-friendly figures will do to help it.

By Brian Dunning
Follow @BrianDunning